This action might not be possible to undo. Are you sure you want to continue?
It is critical for today’s 3G network operators to change their approach to network optimization if they want to remain competitive and profitable. What was effective and acceptable with 2G and 2.5G networks is no longer getting the job done. If network operators want to create a competitive edge while rolling out new services, they must deploy smarter tools and strategies that allow for significant OPEX savings and optimize internal resources. This application note will describe how modern wireless operators can achieve increased network optimization efficiency by overcoming some of the typical shortcomings of a dated optimization strategy inherited from the 2G world. These shortcomings are a result of drive test campaigns and network element counts. They include lack of visibility on uplink activity, poor statistical relevance of data gathered and the high operating costs associated for each campaign. Tektronix has demonstrated time and time again the benefits of implementing a new approach to network optimization – one that integrates the use of protocol-based information with on-demand drive test campaigns. Carriers have the ability to not only to resolve non-effective 3G optimization, but also to realize significant OPEX savings of up to 60 percent. Ultimately, operators are able to successfully roll out new services without quality degradation.
Top 10 Issues Identified by Radio Engineers that are Resolved by OptiMon
1. Ability to clean neighboring cell list - Over-provisioned neighbors - Missing neighbors 2. Ping Pong Handover Detection 3. Propagation Delay
4. Handover Relations 5. Ec/No Analysis 6. RAB Usage and Distribution 7. HSDPA/HSUPA Throughput 8. UE Distance for Services 9. Pilot Pollution 10. Geo-localization of Problems
Application Note | www.tektronixcommunications.com/optimon
But because operators have not had the time to build new competencies. they have no choice but to have their KPIs monitored by professional services. by doing this. while being faced with the constant pressure to reduce operations costs to maximize ROI. in order to cut the time-to-revenues for new 3G infrastructure while simultaneously attending to existing networks. Changes to subscribers’ demographic and usage profiles resulting in an increased need for capacity to cope with the consumption of data-intensive services. and PC cards. Smart Phones (including iPhone applications driving the use of data services beyond just business users).5G networks. Figure 1. MIMO antenna. Unfortunately.Application Note | OptiMon Introduction Today’s wireless communications environment is unlike any we’ve ever experienced. or the need to increase the capacity of the backhaul infrastructure with the provisioning of additional NodeB E1s. There are numerous factors that are occurring and driving the need for more and more optimization activities. As a result. So.5G networks is simply not effective. Examples may include HSPA NodeBs. but may mislead the operator regarding final user experience. Yet another dilemma is the concurrent need to increase the perceived quality of the new services. These factors include: The increasing number of 3G subscribers (ultimately a goal for all carriers). Continuous changes in the surrounding radio environment due to the frequent deployment of new radio sources or changes to city landscapes. These new technology options are pushing networks to the limit in terms of capacity and resource utilization. Using the same approach that worked for 2G and 2. The deployment of new network elements in the RAN. Increase QoS/QoE & Cut Costs 2 . But despite these compelling trigger factors. more wireless carriers rely on vendor professional services teams for network tuning and resources optimization. An increase in network traffic due to the concurrent availability of high speed data services. operators are missing an opportunity to build competences on the more complex 3G radio environment. network operators’ Radio Engineering teams must still contend with 2G/2. data may suit the perspective of the infrastructure vendor. It is also putting operators in a confusing and difficult situation.
at the RNC level. as represented by Figure 2 above). Interface lub lub lub lub luPS luPS Ranking Criteria per IMSI per Handset Type per Cell per NodeB per RNC per Application Service Probablye cause of the problem Specific Subscriber issue Handset Model issue Radio issue E1 link issue PS Core issue Application server issue www. 3G Radio Access is far more complicated than GSM/GPRS technology. As a result. it is easy to get a complete view of performance (call drops.tektronixcommunications.) from the core network (MSC. 80 percent of the problems faced by operators in 3G networks are located in the RAN. Typical Optimization tasks at RNC level It is important to note that just because a wireless service provider has operated its 2G network for quite some time and has established troubleshooting processes. attachment failures. Most of these functions have been transferred to the RNC and they play a central role in the UMTS access network. after analyzing the typical optimization process primarily being used today (i. etc. Sources of Input Data for Optimization: GSM vs.Where most problems arise .e. With the right approach. Figure 3. doesn’t mean that the same process will work well for a 3G network. significant savings and efficiencies can be achieved in each of the tasks represented by blocks in figure 2. Drive testing is a good example of what worked well with 2G. In 3G. but not so well for 3G.By Nature and Design. the MSC is no longer the central point of decision for mobility management. etc.Where the best chance of finding the root cause is possible OptiMon narrows down the possible sources of performance issues: Figure 2. In reality. SGSN). It is also (ironically) the most used type of 3G optimization test (see figure below). not in the Mobile Core Network. UMTS The end-to-end performance is very sensible to the radio quality Start investigation in the RAN . But the most compelling business case can be made for Drive Testing. 2G is Different than 3G What follows are a few examples of just how different these networks are: In a GSM/GPRS network. operators may find that there is lots of room for improvement in terms of effectiveness and efficiency. call management.com/optimon 3 .
These are based on standard routine activities and typical issues. Drive testing can only use limited types of mobiles There are only a few mobile terminals (ME) that are available and compatible with drive test systems. performance indicators and trends. their practical value is limited. the data is collected only from a small fraction of the system coverage area. drive tests are still widely used by 3G UMTS optimization engineers for the following tasks: Network performance assessment Identification of network problems Validation of the effects of applied optimization changes Root problem cause analysis But in fact. The reason lies mainly in the following characteristics of drive test vs. Also. 4 . the data is only collected in a small time interval that usually does not correspond to the system busy hour. In the time domain. it is quite possible that many mobiles may experience a problem that is not sampled during the drive test. if not useless. compared to other test methods. With complete visibility. other test methods: Drive Test data is severely under-sampled This under-sampling occurs both in spatial and time domains. if a problem is detected. Drive testing in 3G networks cannot properly cover new usage profiles In-building (office and home) usage is growing dramatically due the availability of new services and the success of wireless broadband bundles. it is impossible to detect radio coverage problems due to interference in the UL. The first metric to determine is whether or not the time and effort required to investigate trouble tickets are producing any results. This severe under-sampling raises issues regarding statistical validity and repeatability. but also useless when compared with the information that could be gathered using a solution that provides full visibility of network data. all network problems that manifest themselves on other interfaces are usually not visible from the drive test data. the operator will be able to measure the performance of each call from every mobile at the NodeB level. The Unnecessary Deployment and Costs of Drive Test What follows are three specific situations in which the use of drive testing can be misleading. the only activities for which drive testing is well suited for is Problem Root Causes Analysis and Competitive Benchmarking. how certain is the operator that this problem will also occur for other mobile phones? Likewise. A significant example is when the operator must investigate complaints related to frequently dropped calls with large gaps between access areas – probably overshooting. subsequently ignoring the performance of all the other models that subscribers really do use. The operator will also have the ability to spot events. Reproducing customer complaints Many trouble tickets assigned to an optimization department today require drive testing to reproduce a specific behavior or condition related to customer complaints. For example. drive testing is the most expensive test method available for optimization. For example. Drive testing (and in this case the entire campaign) will be not only OPEX-intensive. Drive test data only monitors the Uu interface Therefore. This issue is very hard to detect because of the low statistical sampling that is provided. The typical wireless operator receives approximately 20 trouble tickets per 100. it is easy to understand why drive testing is not relevant when measuring end-to-end performance-of-services. the total cost of ownership is still high. Even in most detailed drive testing scenario.g. Drive testing is very expensive and time consuming In fact. The consequence of this limitation is that drive test measurements can only apply to a few types of mobiles.Application Note | OptiMon Why Drive Testing is not Effective Drive Testing in UMTS networks As discussed earlier. This is because their operation in public transportation vehicles is usually constrained by very limited geographic areas (e.000 subscribers each month. Although automated drive test systems were introduced to reduce the operating cost of drive testing. bus routes). When we consider that end-to-end service performance (end user perceived performance) will vary with each terminal module.
values of -10 dB are required. A histogram that shows distribution of Ec/N0 values will provide the best graphical tool for Down Link quality analysis at cell level. This is significant because the majority of this traffic is business users. A significant 15% of calls can be highlighted as potential overshooting. The report can demonstrate at a glance the distribution of calls by distance. But because of everchanging radio conditions. The left side of the graph represents a small group of users that do not warrant registration on a cell. www. Access Distance Analysis Report The availability of detailed reports and data aggregations allows for faster and more accurate analysis and resolution of customer complaint tickets. but do not yet qualify for cell reselection. In this case.com/optimon 5 . This functionality helps optimization engineers solve issues that might be difficult to identify with only a twohour Drive Test campaign.Let’s consider the example below. the operator would be lucky to identify this problem within a typical two-hour drive test session. normally the first set of customers operators want to keep satisfied. Figure 4. Discrete Distribute Function within Cell Radio Analysis Figure 5. the recommended course of action would be to find a good compromise between both access areas (good/bad). OptiMon’s Cell Radio Analysis application allows for the creation of histograms based on a large base of data. a value of -14dB is sufficient for cell reselection. The Discrete Distribution Function shown in the picture demonstrates how a protocol analysis-based solution can help fill the gaps of Drive Test-based analysis and reduce the time to problem resolution. Experience has shown that decreasing tilt could be an option. The use of the drive test limits the reach of the optimization department to only those areas that are “drive test-accessible”. providing very limited statistical aid.tektronixcommunications. Reducing QoE issues due to Pilot Pollution With the increased availability of HSDPA networks. But for the start of initial registration procedures. Discrete Distribute Function within Cell Radio Analysis Typically. In this example – frequent drop calls with large gaps between access areas – the Access Distance statistics indicate that it is possible to quickly isolate those overshooting instances that would have been otherwise difficult to spot with a methodology based on Drive Test. DL interference is becoming a significant topic with a negative impact on overall QoE of HSDPA traffic.
Traffic can be stored for further analysis. a cluster of NodeBs in one RNC) are monitored and analyzed. Each UE literally becomes a 'probe'. While the total number of affected cell may only be between 0. the impact could in fact be quite painful.. voice quality problems and throughput problems can result in additional trouble tickets costs and an increasing number of potential dissatisfied customers. The 3G Approach to Network Optimization For modern 3G wireless communications. few resources are required. In other words. first impressions matter. QoS issues due to call drops. This method is very effective for both detecting and troubleshooting network problems. So in the end.e. Extending Network visibility with Protocol Analysis-based intelligence 6 . Multiple interfaces that are not visible from the mobiles can be monitored. optimization engineers that are exclusively relying on drive test should quickly consider alternative approaches to prevent quality degradation and revenue loss. Figure 6. This is completely unlike drive test. drive test systems simply cannot measure uplink radio performance. Larger traffic areas can be monitored with one system. protocol-based optimization tools are the ideal choice for assessing network performance and identifying potential network problems. It has been reported that with the launch of new high speed data services in 3G. Typically all calls from a larger area (i.Application Note | OptiMon Up-Link Interference Issues As demonstrated in earlier examples. Some operators realize this but choose to ignore the problem because they feel only a small portion of their cells are affected by up-link interference. IuCP/PS. root cause analysis is very straight forward and time-to-investigate is decreased. Protocol-based optimization solutions (monitoring traffic on Iub. the QoE perceived by the end user during the first attempts to access the services will almost always determine the attitude of that customer when using or buying that service in the future.5%-1%. By recording the data. Iur) are able to obtain a more complete picture of the overall radio access network. Protocol-based optimization tools have the following important advantages: Test results are statistically valid. Once installed. delivering measurement data to the protocol-based optimization system. which demands extensive time and resources.
an effective approach to troubleshooting of data throughput issues has been created. the drive test system can generate calibrated application-level traffic that can be then analyzed by protocol analysis-based solutions on Iub and Iu interfaces. and drive testing was reduced by almost 65 percent. overall team productivity increased (with faster time to trouble resolution). the number of on-demand test cases was reduced due to the troubleshooting capabilities provided by OptiMon. on-demand drive testing was done when needed for troubleshooting. The findings from this trial project were very impressive: Prior to the start of the trial. Drive testing was performed only in cases where the network engineers could not determine the root cause with the protocol-based solution and additional data from the site was required. In summary. Additionally. As a result. In an optimization field trial with a European mobile network operator. But for network performance assessments or problem identification. drive testing was performed regularly according to a predefined schedule. not only was the regularly scheduled drive testing reduced to (almost) zero.com/optimon 7 . www. was installed to observe and analyze a larger geographic area. wireless operators still rely on it for these two features: All measurements gathered by drive test are geographically referenced.Case Study: Drive testing can be reduced by 60% Despite all the limitations of drive test methodology discussed in this application note.tektronixcommunications. but additionally. For example. a protocol-based optimization solution. the following process was implemented: Tektronix’ OptiMon. The end user experience can be assessed by measuring on the Uu interface. protocol-based optimization solutions are the efficient and effective choice. The network engineers used this solution to identify and investigate potential problems. Network operators can instead gain significant efficiency by focusing the use of drive testing only on those areas where it can play off its strengths. During the project.
Drive testing is not effective due to its high operating costs. For Further Information: Tektronix maintains a comprehensive. VoIP and triple play. new test tools. Tektronix. technical briefs and other resources to help engineers working on the cutting edge of technology. Information in this publication supersedes that in all previously published material.Conclusion Although very effective and widely adopted for 2G/2. Specification and price change privileges reserved. By doing this. and foreign patents. and the QoS “traps” created by misleading optimization techniques are avoided. Please visit www. IP and converged multi-service networks. Inc. Tektronix products are covered by U. Tektronix’ experts are available to demonstrate how OptiMon can increase the efficiency and productivity of network optimization departments while simultaneously reducing drive test costs and improving the end user’s Quality of Experience.tektronix. drive test is no longer the “all-in-one” answer to 3G optimization challenges.com/communications Contact Tektronix: Please visit www. About Tektronix Communications: Tektronix Communications provides network operators and equipment manufacturers around the world an unparalleled suite of network diagnostics and management solutions for fixed. issued and pending. OPEX costs are reduced dramatically. including IPTV. relying on protocol-based network optimization solutions for most of the activities and reducing the use of drive test only to specific troubleshooting campaigns and for competitive benchmarking. Wireless operators must redesign their Network Optimization process. mobile. 10/08 | CAW-23083-0 .tektronix. All rights reserved.S. and to the organization itself.5G optimization. TEKTRONIX and TEK are registered trademarks of Tektronix. constantly expanding collection of application notes. trademarks or registered trademarks of their respective companies. fixed mobile convergence. broadband wireless access. IMS. All other trade names referenced are the service marks. Tektronix provides high-level consulting services to assist network operation and engineering departments in the implementation of test method change. as well as its limiting campaign-based approach – its inability to address all the issues that are generated in 3G radio access network.com/communications Phone: 1-800-833-9200 option 1 +1-469-330-4000 Copyright ©. In addition. WiMAX. This comprehensive set of solutions support a range of architectures and applications such as LTE.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.